.png){kind=icon}
返回
.png){kind=icon}
.png){kind=icon}
.png){kind=icon}
.png){kind=icon}
.png){kind=icon}
摘要
En 中文
In this work we design new searchable encryption schemes whose goal is to minimize the
number of cryptographic operations
required to retrieve the result---a dimension mostly overlooked by previous works, yet very important in practice. Our main idea is to utilize compression so as to reduce the size of the plaintext indexes before producing the encrypted searchable indices. Our solution can use any existing Searchable Encryption (SE) scheme as a black-box and any combination of lossless compression algorithms, without compromising security.
The efficiency of our schemes varies based on the leakage exposed by the underlying application. For instance, for private keyword search (more leakage), we demonstrate up to 188× savings in search time, while for database search (less leakage) our savings are up to 62×.
The power of our approach is better manifested when combined with more secure, yet less practical, cryptographic tools, such as Oblivious Random Access Memory (ORAM). In particular while ORAM is known to be prohibitively expensive for large-scale applications, we show that our compress-first-ORAM-next approach allows significant more efficient index search time, reducing the time for executing a query with result of size more than one million tuples from approximately 21 hours to 20 minutes.
期刊
暂无期刊信息
机构
暂无机构信息
引用论文
暂无论文信息